This proposal addresses two important areas in the pathogenesis of HIV-1 infection: (1) What cellular pathways contribute to viral replication, particularly in the early "subclinical" stages of disease? What cells are infected at mucosal surfaces, how is virus transmitted to T cells, and why does viral replication proceed chronically at substantial levels even in asymptomatic individuals? (2) What are the mechanisms whereby CD4+ T cells are lost, particularly T cells that mediate immunologic memory? Why is the loss of CD4+ T cells, like viral replication, apparently a chronic process such that HIV-1 infected individuals may lose > 109 CD4+ T cells each day? The overall hypothesis is that dendritic cells [DCs] play an important role in both areas of pathogenesis, because of the capacity of DCs to express CD4 and to interact with CD4+ memory T cells in mucosal and extravascular sites. To model mucosal leukocytes, the PI has studied DCs and T cells from skin. He finds that small amounts of virus in DCs can, in the presence of CD4+ T cells, lead to HIV-1 replication and apoptotic T cell death in the absence of exogenous stimuli. He have now visualized productive infection of DCs in mucosa of infected patients early in the course of their disease. He will now pursue 3 sets of questions on DC function in situ and in vitro: (1) Are productively infected DCs [identified with a panel of new antibodies to DCs] present in several different mucosal tissues, and in cells isolated from afferent lymph? (2) What are the viral and cellular requirements for vigorous replication of HIV-1 in mixtures of skin derived DCs and T cells? He will emphasize fresh viral isolates from asymptomatic patients, and a group of adhesion and costimulatory molecules that mediate the DC-T cell interaction. (3) What are the viral and cellular requirements for the apoptotic killing of CD4+ memory T cells? He will pursue preliminary findings for bystander killing of noninfected T cells.